Monday, March 23, 2020
BAD MATH ALERT: No, the seasonal flu doesn’t have a 10% fatality rate
Let me spare you a little back of the envelope calculation. We alreacdy know that flu caused around 35,000 deaths during the last flu season. We can expect COV19 to produce at least twice that or around 70,000 minimum and conservatively around 100,000 dead for the USA.
These numbers are still low when put in perspective in terms of actual potential. I am optimistic that the steps taken so far will work to mitigate along with rapidly improving weather.care of soke sort
However, do expect a huge number of folks needing care of some sort, way beyond what is typical. This induces a two week fever hitting 103 which is no joke at all. You are sick and someone will have to keep you supplied with ample water and apple juice and ginger ale and all that to help. The hospitals will struggle with the worst cases, while the rest of us have to help each other.
This will be no joke. Testing is now settling down and we are still in the exponential phase of case acquisition. No fun to watch.
BAD MATH ALERT: No, the seasonal flu doesn’t have a 10% fatality rate… it’s actually less than 0.1%… here’s the math
Tags: badhealth, badmedicine, badscience, Collapse, coronavirus, infections, outbreak, pandemic, The Gateway Pundit
The article essentially claims that the seasonal flu kills 10% of those it infects, but the math approach is incorrect
Deaths lag several weeks behind confirmed infections due to the “dying lag time”
Look around… if the seasonal flu really killed 10% of infected people, you would be encountering a steady stream of dead bodies every winter
The “unknown denominator” fallacy pushed by Rush Limbaugh
So how do we know anything at all about the case fatality rate?
Call me Mr. Semmelweis
Monday, March 16, 2020 by: Mike Adams
One of the serious problems we are dealing with right now in our world is the fact that pro-Trump publishers and pundits have largely decided the coronavirus is a hoax because acknowledging the existence of the pandemic might hurt Trump, they figure. This delusional thinking is summed up by the new mantra of right-wing denialism: “It’s no worse than the flu.”
Rush Limbaugh has been touting that line for weeks, oblivious to the mathematics of epidemiology and viral transmission. Even celebrity doctors like Dr. Drew are recklessly downplaying the severity of this outbreak, following several weeks of President Trump initially claiming the virus was no big deal and would just magically disappear. (He first said that when America had 15 infections, by the way. We now have 4,600+. Thankfully, Trump quickly changed his tune and began to get more serious about it.)
But the worst case of bad information we’ve seen so far happens to come from our friends at The Gateway Pundit, where a column by Joe Hoft (the brother of Jim Hoft, the publisher) has written an article that essentially claims the case fatality rate for the regular flu is 10%.
We say this with respect for the monumental effort that TGP is making to survive and prosper in an age of extreme censorship of pro-Trump voices, but the claim by Joe Hoft overestimates the actual case fatality rate of the flu by over 100 times and achieves the dissemination of truly misleading information that may persuade many conservatives to downplay the severity of this virus to their own demise. If Trump loses in November, it will likely be because so many conservatives died from following bad coronavirus information, enabling the Democrats to drop-kick a living, walking zombie named Joe Biden across the finish line.
The Gateway Pundit, by the way, shares our fate of being banned by every left-wing tech platform for the crime of being effective during the 2015 – 2016 election season at helping get Trump elected. Since then, TGP has exposed the lunacy of left-wing politicians on a daily basis, earning a high amount of readership among Trump supporters and conservatives in general. We’re fans of TGP’s previous work, which is why we bring this up: It’s important for TGP to get the math right on this issue. And so far, the truly bad math on their article doesn’t rise to the level of reporting we would expect from TGP, which is why I’m bringing it up here, perhaps at risk of getting criticized myself by the many TGP fans. Nonetheless, I’ll stick my neck out on this issue, since the real math is quite clear, and I’m pretty damn good at math, being that I’m a lab scientist who does this sort of math every day.
In fact, my pandemic projection model has been shockingly accurate so far at predicting the numbers of infections and deaths. I’ll cover that in another post.
With apologies to Joe Hoft if any of this seems unjustifiably harsh, here’s how he arrives at his 10% figure for the seasonal flu:
Citing this CDC page, Hoft’s article examines the number of current deaths this year from the flu, which the CDC estimates to be 22,000. Notably, this number is an estimate by the CDC, not proof that 22,000 deaths were confirmed to have been caused by seasonal influenza.
Joe Hoft then divides this 22,000 estimate by the “No. of positive specimens” reported by the CDC under the column heading, “Data Cumulative since September 29, 2019 (week 40)” which gives the number of 222,552 positive specimens having been tested so far. Importantly, this is talking about actual tests, not a projection or estimate of the total number infected.
From this, he calculates 22,000 / 222,000 and concludes that the seasonal flu kills 10% of patients who get it.
He then goes on to reach the completely wrong conclusion, stating, “Actual results for the coronavirus are lower than the flu.”
There’s just one problem in all this. The comparison is taking the CDC’s estimated number of 22,000 deaths and comparing it to the actual number of confirmed specimen tests. In other words, he’s doing the math like this:
Fatality rate = Estimated deaths / Actual confirmed tests of infected
This is a critical error, and it renders the 10% answer completely incorrect. The CDC even says on its page that the “22,000 deaths from flu” is a “CDC estimate,” not an actual number of confirmed cases.
The correct way to run this math would be more like this:
Estimated / Estimated
Actual / Actual
If you use estimated numbers for both the numerator and denominator, you get this:
Fatality rate = 22,000 (estimated) / 36 million (estimated)
And what does that answer come to?
That’s not “six percent.” It’s actually “six-tenths of one-tenth of one percent,” or stated another way, “Six-one-hundredths of one percent.”
Or stated another way, it means that for every 1,636 people who are infected, ONE person dies from the seasonal flu (in this season).
In other words, the case fatality rate is nowhere near 10% as concluded by Joe Hoft. He overstates the case fatality rate by more than 100 times, which is of course two orders of magnitude. That’s a huge error, and it misleads people who can’t recognize the flaw in the math because, of course, we live in a country where nobody can do math, sadly.
When looking at the coronavirus case fatality rate, for every 1000 people who are infected, somewhere from 20 to 60 people die. (Roughly 2% – 6% case fatality rate.)
Even if you take the lowest case fatality rate of just 2%, that’s 33 times higher than the seasonal flu fatality rate of 0.06% in this case.
So no, the seasonal flu death rate is not higher than the coronavirus death rate. It’s not even close. It’s not even in the same ballpark.
For the sake of keeping the numbers relatively easy to handle, we typically use 0.1% as the case fatality rate for the seasonal flu, even though in many years the actual case fatality rate is significantly lower such as 0.06%. But even if we use 0.1%, it means the coronavirus is at least 20 times more deadly than the flu and might be more like 57 times more deadly, when it achieves a 5.7% case fatality rate.
We also know this fatality rate skyrockets to something much closer to 15% when hospitals get overrun and first-line medical care is no longer available to the flood of new patients. In those cases, the fatality rate leaves the seasonal flu in the dust and leads to nations like Italy suffering near-collapse conditions of their entire national health care infrastructure. (Which the flu doesn’t achieve, of course, since the seasonal flu kills almost no one by comparison.)
There’s another issue in all this that needs to be considered: There’s a lag time between new infections and deaths. This lag time is many weeks in duration, given that it takes 1-2 weeks to start to show symptoms, then another week or so to end up in the hospital, and then roughly another 2-3 weeks to expire, typically. So the deaths lag behind the new infections by several weeks, which means that when new infections explode exponentially in a given region — such as NYC — the deaths follow a trailing curve that hits you roughly 4-5 weeks later (which is exactly what we saw in Italy, which also downplayed the severity of all this until their hospital system was brought to the brink of collapse).
So when dealing with the coronavirus, you can’t just take the number of deaths from today and divide by the number of infections from today. That would give you the wrong answer. You actually need to take the deaths today vs. the number of infections from 4-5 weeks ago. That’s your real death rate (technically, the “case fatality” rate).
I find it stunning that almost no one is doing the math accurately on this. Chris Martenson from Peak Prosperity is nailing it, by the way, and I don’t know his politics, but that’s irrelevant right now. He’s getting the math right, which is far more important than your political beliefs at this particular moment in human history.
Getting back to the article by Joe Hoft, where Joe is correct is that the CDC doesn’t actually test 36 million people every year for the flu. That’s a projection, and part of that projection is no doubt an attempt to push flu vaccine propaganda. However, even from our own experience, we know that plenty of people around us get sick during the flu season. You don’t have to have the CDC tell you that, roughly speaking, at least 1 out of 10 of the people you know gets sick each flu season, on average. With a national population of 327 million, 1 in 10 getting the flu each year is about 33 million people, which is right about where the CDC estimates land.
The CDC publishes its estimates at this CDC.gov web page:
There, you’ll see that the estimates for the 2018 – 2019 flu season calculate that roughly 35.5 million Americans got infected with the flu, 490,561 people required hospitalization, and 34,157 people died.
34,000 deaths out of 35.5 million infected is almost exactly a 0.1% case fatality rate, which is 100 times lower than Joe Hoft’s analysis from TGP.
Naturally, the CDC doesn’t really run around and test 35.5 million people for the flu. In fact, we’ve come to learn recently that the CDC is apparently incapable of testing anyone for anything, especially when the coronavirus is the subject at hand. But we don’t have to trust the CDC numbers in the first place. We can derive some fairly good answers out of our own experience.
How many times have you caught the winter flu in your lifetime? Ten times? Twenty? Forty?
If Joe Hoft’s analysis is correct, you would have roughly a 10% chance of dying each time you got the flu. So if you had the flu ten times, your odds of being killed by at least one of those infections would be described by this simple equation: 1 – (0.9^10), expressed as a percentage.
Over ten flu infections, that means there is a 65% chance you would have been killed already.
Look around at your family and friends. Are two-thirds of them missing after ten bouts of the regular flu?
Unless your family members happen to have dirt on the Clintons, the answer is probably not.
We see people getting the flu all the time: family members, co-workers, spouses, partners and so on. We don’t see them dropping dead 10% of the time they get it. The flu isn’t very deadly at all, which is why we don’t shut down society to fight the regular flu. (It’s also something that our bodies have encountered before, unlike the novel coronavirus, which is called “novel” because it’s new to human biology, and that’s part of what makes it so deadly.)
We accept the 35,000 – 65,000 total deaths every year from the flu as just a normal part of the economy. But if those numbers were 57 times higher — which seems to be where the coronavirus numbers are — then you have to take aggressive actions to halt millions of people from perishing. Hence the social distancing protocols being recommended by Trump and Fauci.
Here’s another math issue that Rush Limbaugh seems to stumble with, since Rush has no formal education in the sciences whatsoever. He argues that since testing is not widespread yet in the United States, we know for certain how many people died from the coronavirus but we don’t know how many had it but didn’t die. Thus, says Rush, we know with certainty the numerator of this equation, but not the denominator:
# of people who died / # of people who had it
However, his explanation is a logical fallacy. In truth, because we aren’t doing much testing, we also don’t know how many have died from it, because as we saw from China, perhaps 100,000 people or more were said by the communist regime to have died from “pneumonia” instead of the coronavirus.
Even in the United States, nobody is testing dead people to find out whether they had the virus. Thus, in truth, we don’t even know the numerator of that fraction described above.
So Rush Limbaugh’s logic is faulty. It assumes we have an accurate numerator, which we don’t. Rush completely fails to point this out since he’s following a political agenda on this, not legitimate mathematics or science.
Given all this, you might argue how do we know anything at all about the case fatality rate of the coronavirus? If the numerator is unknown and the denominator is unknown, then how can we calculate the final case fatality rate with any sense of certainty?
For starters, when Iran is digging mass graves so large that they can be seen from orbit, you know you have a death problem on your hands.
Furthermore, we can look at China’s response to this and realize that China saw the fatalities as being so high, it risked financial ruin and a political collapse to lock down its society and halt the spread. That’s not something China would do except as a last-ditch resort to prevent a far worse apocalypse from taking place. China, after all, isn’t in the self-destruct business, at least not on purpose.
You can also look at the hospitals of Northern Italy and how they are being overrun with the dying and dead. That’s not “just the flu.” It’s something far, far worse.
Alternatively, if you wanted to try to convince yourself the coronavirus is no big deal, you could invoke the conspiracy theory that claims Italy, Iran, China, Taiwan, Australia, Singapore, South Korea, France, Germany and over a hundred other countries are all just faking it for some inexplicable reason.
But that would be insane, not to mention impossible to coordinate. Iran, China, Taiwan and France can hardly sit at the same table at an international delegation, much less coordinate some grand global “fake death” conspiracy to pull off some sort of bizarre world theater for some unknown reason.
So unless you believe all these countries are faking it, you know this is far worse than the flu. And anyone saying the flu is more deadly than the coronavirus, flatly stated, is wildly misinformed.
Here are some more general thoughts that aren’t specific to Hoft’s article but what I’m experiencing in general when trying to educate conservatives about math and science.
In this coronavirus fiasco, there are days where I feel a bit of what it must have been like to be the first person on the planet to have discovered the germ theory and then be tasked with explaining germs to everybody else, all of whom are completely clueless and have dismissed germs as some sort of “invisible conspiracy theory” since they can’t see germs with their own eyes.
Call me Mr. Semmelweis from now on, and realize that it’s torture to be able to do basic math and science when living in a world full of people who can’t.
After all, it’s hard for us to mock Brian Williams for being off by six orders of magnitude on the Bloomberg funding question, if we are also off by two or more orders of magnitude in our own reporting. (For the record, Brian Williams claimed that $500 million in Bloomberg money could have given 327 million Americans $1 million each. In truth, Bloomberg would have had to come up with $327 trillion to achieve that, and $1 trillion is 10^6 times $1 million, which is where the “six orders of magnitude” comes from, but then again, nobody can do math in America anymore so I guess numbers don’t mean anything, so this whole exercise is pointless.)
I will share with you one final number that you need to become acquainted with, and that number is ZERO.
Because ZERO is what many of Wall Street’s “darling” corporations are actually worth once you deduct debts and liabilities from assets and revenues. And ZERO is what many are headed toward, even with all the Fed pumping and insane bailouts that are nothing more than government-organized looting of taxpayers’ money.
To understand what’s coming, ZERO is the most important number you need to remember, for ultimately that’s what the US dollar will be worth, too, once all the bailouts and fiat currency printing and bankrupt spending spree is over and done with.
As in, Zero Hedge, which is one of the few independent publishers that’s actually doing some damn fine reporting on this coronavirus situation. Perhaps you haven’t noticed their site slogan in a while, but it reads, “On a long enough timeline the survival rate for everyone drops to zero.”
They’re correct, but some people are accelerating the timeline by being idiots when it comes to dealing with the coronavirus.
Try not to be among the ones who die of ignorance because they failed to recognize the severity of this pandemic before it was too late.
And for God’s sake, enough with the toilet paper, okay?